MAPREDUCE-7101. Add config parameter to allow JHS to alway scan user dir irrespective of modTime. (Thomas Marquardt via asuresh)

Author: xslogic

hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/main/java/org/apache/hadoop/mapreduce/v2/jobhistory/JHAdminConfig.java

@@ -61,8 +61,13 @@ public class JHAdminConfig {
     MR_HISTORY_PREFIX + "cleaner.interval-ms";
   public static final long DEFAULT_MR_HISTORY_CLEANER_INTERVAL_MS = 
     1 * 24 * 60 * 60 * 1000l; //1 day
-  
-  
+
+  /** Always scan user dir, irrespective of dir modification time.*/
+  public static final String MR_HISTORY_ALWAYS_SCAN_USER_DIR =
+      MR_HISTORY_PREFIX + "always-scan-user-dir";
+  public static final boolean DEFAULT_MR_HISTORY_ALWAYS_SCAN_USER_DIR =
+      false;
+
   /** The number of threads to handle client API requests.*/
   public static final String MR_HISTORY_CLIENT_THREAD_COUNT = 
     MR_HISTORY_PREFIX + "client.thread-count";

hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/resources/mapred-default.xml

@@ -1774,6 +1774,15 @@
   </description>
 </property>
 
+<property>
+  <name>mapreduce.jobhistory.always-scan-user-dir</name>
+  <value>false</value>
+  <description>Some Cloud FileSystems do not currently update the
+  modification time of directories. To support these filesystems, this
+  configuration value should be set to 'true'.
+  </description>
+</property>
+
 <property>
   <name>mapreduce.jobhistory.done-dir</name>
   <value>${yarn.app.mapreduce.am.staging-dir}/history/done</value>

hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-hs/src/main/java/org/apache/hadoop/mapreduce/v2/hs/HistoryFileManager.java

@@ -324,7 +324,13 @@ public synchronized void scanIfNeeded(FileStatus fs) {
       // so we need to have additional check.
       // Note: modTime (X second Y millisecond) could be casted to X second or
       // X+1 second.
-      if (modTime != newModTime
+      // MAPREDUCE-7101: Some Cloud FileSystems do not currently update the
+      // modification time of directories. For these, we scan every time if
+      // the 'alwaysScan' is true.
+      boolean alwaysScan = conf.getBoolean(
+          JHAdminConfig.MR_HISTORY_ALWAYS_SCAN_USER_DIR,
+          JHAdminConfig.DEFAULT_MR_HISTORY_ALWAYS_SCAN_USER_DIR);
+      if (alwaysScan || modTime != newModTime
           || (scanTime/1000) == (modTime/1000)
           || (scanTime/1000 + 1) == (modTime/1000)) {
         // reset scanTime before scanning happens